VEHICLE-HEALTH INTELLIGENCE PLATFORM

Know the vehicle.
Understand its health.

KryLocke-vHIP turns raw OBD, CAN, diagnostic and historical vehicle data into a living health picture—across ICE, hybrid and electric platforms. It is no longer a concept: the app reads live vehicles today.

LiveReading vehicles now
Read-onlySafety by design
Evidence-firstAI explains, never commands
High-performance vehicle rendered for KryLocke-vHIP
34PIDS DECODED
LIVEOBD / CAN
AIINSIGHT
DIAGNOSE•ANALYZE•MONITOR•ASSESS•COMPARE•EXPLAIN•PROTECT
FIRST LIVE SESSION

It has met a real car.

A 2010 Infiniti G37 at 100,797 miles, read end to end over a vLinker MS adapter. Not a simulation and not a demo file — the figures below are what the vehicle actually returned.

46 / 34PIDs supported / decoded
7 / 7Readiness monitors read
13.3 VModule supply measured
2Fault codes confirmed

Protocol ISO 15765-4 CAN, 11-bit, 500 kbaud · responding module 7E8 · ELM327 v2.3 adapter identified and verified

Vehicle VIN read and decoded from the vehicle itself, matched against the garage profile

Findings P0300 random/multiple cylinder misfire and P0021 camshaft timing over-advance — each returned as stored, pending and permanent, which is the distinction that tells a technician whether a repair actually held

FROM SCANNER TO INTELLIGENCE

Not another code reader.

KryLocke-vHIP is designed as a vehicle-health record: current state, historical baseline, diagnostic evidence, trends, and AI-assisted interpretation in one place.

Diagnose icon01

Diagnose

Read standardized parameters, fault codes, freeze-frame data and readiness monitors across the modules that answer—without burying the driver in raw codes.

Analyze icon02

Analyze

Normalize signals into health domains, compare them to thresholds, and look for relationships across multiple measurements rather than one code at a time.

Monitor icon03

Monitor

Every finished session is filed against its vehicle automatically and compared with the one before it, so a code that came back reads differently from a code that never left.

AI insights icon04

Explain

Give AI structured evidence—not uncontrolled access to the car—so probable causes, confidence and next tests stay grounded in what was measured.

Look & Listen: photo, video and sound filed per vehicleSession history with before-and-after comparisonSelectable scan depthWorks with no signal
Vehicle side profile
POWERTRAIN94Stable
12V SYSTEM13.3VNormal
NETWORK98Healthy
VEHICLE HEALTH SCORE

A health model that adapts to the vehicle.

ICE and electric vehicles do not expose the same data. KryLocke-vHIP uses the vehicle profile to activate only the health domains that matter, and grades every value by how it was established.

Motor / PowertrainHV BatteryChargingLow VoltageThermalBrakingNetworkSafety Systems
92/100
Vehicle health: Excellent

Current measurements are stable. Two trends are being watched against this vehicle's historical baseline.

AUCTION ASSESSMENT

Decide on a lot in the ten minutes you get beside it.

A buyer at a salvage or dealer auction has a few minutes, bad light and no signal. KryLocke-vHIP turns that into a structured record: the facts, a guided photo set, five scores, the notices worth acting on, and the most you can pay and still make money.

RebuildCan it realistically go back on the road?
SalabilityOnce fixed, how readily does it sell?
CostHow favourable is the money, all in?
BiddingWill you win it at a price that still works?
ConfidenceHow much rests on fact rather than a blank?

Confidence is the score that keeps the other four honest.

It is the only dimension that falls when something has not been checked. Without it, a scorecard built on three facts and nine blanks looks exactly like one built on twelve—same five numbers, same clean average, no hint that most of it was inference. Printed beside the others, the thinness of an assessment is visible at the same glance as its conclusion.

Every score is tappable down to the individual facts that moved it. A number nobody can account for is not a score; it is a hunch with a digit attached.

Plain average of all five — no weighting to hide a weak inputScored on-device, offline, deterministicallyVIN captured by camera or typed, check-digit verified
VIN FIRST · THEN 12 REQUIRED + 4 CLOSE-UPS

Type it or photograph the plate. The VIN sits at the top of both the lot sheet and the walkaround, read on device from the plate's barcode or its printed characters, and checked against its own check digit. Nothing is accepted until you confirm it against the plate — a misread VIN does not announce itself, it quietly becomes a different car.

  • Front
  • Front left corner
  • Left side
  • Rear left corner
  • Rear
  • Rear right corner
  • Right side
  • Front right corner
  • Engine bay
  • Interior from the door
  • Driver's seat
  • Center console
  • Bottom rear
  • Wheel
  • Battery
  • Fuse box

The same frames, in the same order, for every lot — so two cars can be compared and a missing shot means "nobody looked" instead of nothing at all. Guidance for each frame stays on screen while the camera is up.

MAXIMUM BID

Solved as resale, less your margin, less repairs, fees and transport — printed with its own arithmetic so it can be checked rather than trusted. When there is no resale basis it shows nothing rather than a guess: a ceiling computed from nothing is the most dangerous number a screen like this could print.

NOTICES WITH SOMETHING TO DO

Title brands, flood and fire, deployed airbags, suspected structure, high-voltage packs, missing keys, uncertified odometers. Every notice above a passing note carries the one thing to go verify, and links the free official lookups — NHTSA recalls, NICB VINCheck, NMVTIS — all of which need the VIN the app just captured.

MARKET RESEARCH, LABELLED

With a connection, an AI pass estimates retail and wholesale ranges, comparable salvage money and where the repair cost actually goes. Those figures arrive stamped as estimates, and they cost confidence — because nobody has verified them against a sale.

EVIDENCE-FIRST AI

AI that explains the vehicle instead of guessing at it.

The diagnostic core interprets known values first. AI then ranks probable causes, cites supporting evidence, identifies contradictions, and recommends the next read-only tests.

Bring your own key — OpenAI, Anthropic or Google — held in the device Keychain and never in the app's source. KryLocke asks each provider what models the key can actually reach and prefers the small, fast one, because this is schema-constrained reading of structured evidence, not an essay competition. When a heavier model is worth the money, Second opinion runs the larger one alongside and shows where the two disagree. Two readings are never silently merged into one confident answer.

FACTOBSERVATIONINFERENCERECOMMENDATION
KryLocke Insight Engine

01Finding Low-voltage instability detected during three recent sessions.

02Evidence Module voltage dropped from 13.8V to 11.7V during start sequence.

03Correlation Communication DTCs appeared in four modules during the same voltage event.

04Assessment Battery or DC-DC supply issue is more likely than four independent module failures.

05Next test Record 12V minimum voltage during start and verify charging output under load.

91% confidence
WORK THE SHOP CAN USE

From the lane to the desk, without retyping it.

A diagnostic that stops at a phone screen has not finished. KryLocke-vHIP produces the artefacts a service department already works from, and connects to DriveBridge DealerOS so an assessment reaches the desks that act on it.

Intake

Stock-in: identity, VIN, odometer, lot and acquisition facts, with the photo set attached.

Service

Recon worksheet: the damage, the notices worth acting on, and the repair estimate as recommended lines.

Title

Brand, issuing state, odometer certification, and the paperwork that has to clear before retail.

Sales

Pricing and disclosure: the valuation with its basis, the branded-title disclosure, and the walkaround.

Reports a service advisor can paste.

The repair-order block is written in the department's own three C's — concern, cause, correction — with every correction marked recommended and always in technician wording, because a repair order is a technician document. CSV and JSON exports carry the same content for importers, and the provenance line rides on every row: a spreadsheet gets filtered and re-sorted, and no row of this export should ever be read without the limits it was produced under.

Nothing is written into a system we have not been invited into.

CDK, Reynolds, Tekion and Cox all gate write access behind certified partner programmes. KryLocke records which system a shop runs and shapes its export accordingly — it does not claim a live link to any of them. DriveBridge DealerOS is different: both ends are ours, so that one is a real two-way connection, and the app ships ready for it with the server side still being built.

HOW IT CONNECTS TO DRIVEBRIDGE
01Read

The adapter reads the vehicle, read-only, on the lot or in the bay.

02Assess

KryLocke-vHIP builds the health picture, scores and labelled estimates on the device.

03Hand off

The assessment is packaged for DriveBridge DealerOS with its provenance attached.

04Route

DriveBridge places each piece in the right lane: Intake, Service, Title or Sales.

05Return

Status flows back, so the app shows what the desk has acted on. Two-way, because both ends are ours.

The app ships ready for this connection; DriveBridge's server side is still being built. Until it is live, the same content leaves as CSV and JSON exports.

WHY IT HELPS A DEALERSHIP

Support that shows up at every desk.

No retyping

One read feeds stock-in, the recon worksheet, title and pricing. Fewer hand-keyed VINs, fewer transcription errors.

Faster recon

Service sees stored, pending and permanent codes with recommended lines before the car reaches the bay.

Advisor-ready

Repair-order blocks arrive in concern, cause, correction, written in technician wording.

Honest disclosure

Sales gets the valuation with its basis and branded-title facts, so disclosures match the evidence.

Traceable decisions

Every row carries its provenance, so managers can see what was measured and what was estimated.

Safe on the lot

Read-only by design. Nothing is written to the vehicle, and nothing goes into a DMS uninvited.

MODULAR BY DESIGN

Start with vLinker. Grow into a diagnostic platform.

The first iOS release connects through the vLinker MS. The architecture remains interface-neutral so future BLE, Wi-Fi, replay, CAN-FD and J2534 bridge support can be added without rebuilding the app.

Vehicle
ICE / EV / Hybrid
→
VehicleInterface
vLinker MS first
→
Transport
OBD / ISO-TP
→
Knowledge
PID / DTC / Profile
→
HealthAnalyzer
Rules + Trends
→
AIProvider
Explain + Compare
Auction Assessment
Facts + walkaround
→
Scoring
Five scores, on-device
→
Market research
Estimates, labelled
→
DriveBridge
Intake · Service · Title · Sales
FUTURE J2534

A later desktop/network bridge can connect professional SAE J2534 pass-through hardware to KryLocke-vHIP while preserving the same iOS application architecture.

WHAT IT WILL NOT DO

The limits are part of the design.

Every one of these is a decision, not a gap waiting to be filled.

Read-only on the vehicle. The diagnostic path observes. It does not program modules, write memory, or touch immobiliser or security functions.
AI never commands the car. It receives structured evidence and returns an interpretation. Every action stays with the human who initiated it.
An AI figure stays an AI figure. Research-derived numbers are labelled as estimates everywhere they appear, travel labelled into any export, and never quietly become verified vehicle knowledge.
No live auction data. Copart, IAAI and Manheim publish no API for an app like this and their terms forbid automated access. Bids are figures a person observed, logged with the time they saw them — never scraped, never stale and presented as current.
No uninvited writes into a dealer system. Third-party DMS integration means a certified partner programme, and until there is one, KryLocke produces exports rather than pretending to a connection.
V1 DEVELOPMENT ROADMAP

Build useful capability in measurable checkpoints.

Checkpoints are only marked proven once a real vehicle has returned the data.

CP-01

Bluetooth

Discover and reliably connect to vLinker MS.

Proven on vehicle
CP-02

Adapter

Initialize adapter and validate commands.

Proven on vehicle
CP-03

Vehicle

Detect protocol and retrieve VIN.

Proven on vehicle
CP-04

Live Data

Decode standardized SAE parameters.

Proven on vehicle
CP-05

DTC

Read and interpret diagnostic trouble codes.

Proven on vehicle
CP-06

History

Sessions filed per vehicle and compared with the one before.

Built
CP-08

ISO-TP

Reconstruct multi-frame messages reliably.

Built
CP-10

AI

Explain structured evidence without controlling the vehicle.

Built
KryLocke shield
KRYLOCKE-vHIP v1

Turn every diagnostic session into vehicle intelligence.

Reading live vehicles today, assessing auction lots offline, and built to grow into a living health record for modern vehicles.

Pilot inquiry